ABSTRACT
BACKGROUND: We assessed the contribution of dedicated computed tomographic (CT) interpretation to the accuracy of positron emission tomography (PET) plus CT in imaging patients with suspected primary or metastatic colorectal carcinoma. METHODS: One hundred PET/CT scans in 90 consecutive patients were evaluated retrospectively. Imaging was performed on a GE Discovery LS PET/CT scanner. PET images were obtained from the skull base through the midthigh after intravenous administration of 15 to 20 mCi of [(18)F] fluorine-18-fluoro-2-deoxyglucose. Noncontrast axial CT images were obtained at the same anatomic locations, with 140 kV, 80 mA, 0.8 s/CT rotation, a pitch of 6, and a table speed of 22.5 mm/s. The CT component of the PET/CT study was reviewed independently by consensus of two blinded readers. Scans were evaluated for the presence of primary disease, local recurrence, and distant metastases. Results were compared with the PET/CT report. The gold standard was clinical and imaging follow-up for at least 6 months, surgery, or biopsy. RESULTS: The study included 40 males and 50 females, with a mean age of 63 years (range, 31-92 years). The indications for the examination were to evaluate for recurrence of colorectal cancer in 83 cases, determine disease spread in 15 cases, and evaluate for possible primary malignancy in two cases with rising carcinoembryonic antigen. Sensitivity, specificity, and accuracy of the PET/CT report and of the combined PET/CT with dedicated CT interpretation were 0.914, 0.633, and 0.830 and 0.986, 1.000, and 0.980, respectively. The difference between PET/CT and the combined PET/CT with dedicated CT interpretation with respect to accuracy was statistically significant (p < 0.05). CONCLUSION: The CT portion of PET/CT provides valuable anatomic and pathologic information to the functional information provided by PET and helps improve the overall accuracy of the combined study.
